Article depolarization recommendation method and system

Abstract

The invention belongs to the related technical field of information recommendation, and discloses an article depolarization recommendation method and system, and the method comprises the steps: obtaining a user interaction historical sequence and a user article interaction graph, generating text embedded representations of a user and an article, optimizing the embedded representation of each node through comparative learning, and then carrying out graph convolution feature extraction, the method comprises the steps of obtaining a text embedding representation of a target user, obtaining a final graph embedding representation of a node, carrying out feature fusion on the text embedding representation of the target user and the final graph embedding representation to obtain a final embedding representation of the target user, calculating the similarity between the final embedding representation of the target user and the text embedding representation of an article, and determining the recommendation priority of the article according to the similarity. Through the method, the popularity deviation of the recommendation system can be removed, the articles conforming to the real interests of the user are recommended, and the use experience of the user is optimized.

Description

Technical Field

[0001] The present invention belongs to the technical field related to information recommendation, and more specifically, relates to an item debiasing recommendation method and a system thereof. Background Art

[0002] As an important branch of computer information technology, recommendation systems have become an indispensable part of Internet technology in the context of the explosive growth of Internet information. The amount of information on various social media network platforms is increasing day by day, which leads to users facing the challenge of information overload. Information overload makes it difficult for users to filter out useful content from massive amounts of information, but recommendation systems can effectively cope with this challenge.

[0003] With the widespread application of recommendation systems, more and more users have accessed them, and many problems have gradually been exposed. One of the problems is that the current recommendation system often cannot accurately recommend things that meet users' real interests due to the influence of popularity bias. The recommendation system without popularity bias aims to reduce or eliminate the bias in the recommendation results caused by the popularity of items, ensuring that the recommendation list can be more fair and diverse and meet the real interests and needs of users.

[0004] Currently, the removal of popularity bias in recommendation systems generally involves modifying the loss function or post-processing the recommendation results. This method does not consider the balance between debiasing performance and recommendation performance, resulting in the loss of recommendation performance while removing the bias. Summary of the invention

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides an item debiasing recommendation method and system thereof, which aims to remove popularity bias and recommend items that meet the user's real interests.

[0006] To achieve the above object, the present invention provides an item debiasing recommendation method, which comprises:

[0007] Obtain a user interaction history sequence and a user-item interaction graph, wherein the user interaction history sequence includes the user's historical interaction items, and the user-item interaction graph is an intersection adjacency matrix of users and items; concatenate the attribute text representations of all items in the same user interaction history sequence, perform feature extraction and fusion, and obtain a text embedding representation of the corresponding user; perform feature extraction and fusion on the attribute text representation of the same item, and obtain a text embedding representation of the corresponding item; the attribute text of the item includes popularity-related attributes and non-popularity-related attributes;

[0008] Using the text embedding representation of the user and the text embedding representation of the item as the initial node embedding representation of the corresponding node in the user-item interaction graph; traversing all nodes in the user-item interaction graph, dividing the directly connected nodes of the current node into two sets, respectively aggregating the nodes in each set, using one of the aggregation results as the embedding representation of the current node, using the other aggregation result as the positive sample of the current node, randomly selecting positive samples of multiple nodes from nodes other than the directly connected nodes as negative samples of the current node, and optimizing the embedding representation of each node through comparative learning;

[0009] Perform graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the nodes;

[0010] The text embedding representation of the target user and the final graph embedding representation are feature-fused to obtain the final embedding representation of the target user;

[0011] Calculate the similarity between the final embedding representation of the target user and the text embedding representation of the item, and determine the recommendation priority of the item based on the similarity. The higher the similarity, the higher the recommendation priority.

[0012] In some embodiments, obtaining the user interaction history sequence and the user-item interaction graph includes:

[0013] After removing items and users whose interaction numbers are lower than a preset requirement, the user interaction history sequence and the user-item interaction graph are generated according to the interaction records.

[0014] In some embodiments, the feature extraction and fusion comprises:

[0015] First, use the Transformer encoder to extract features from the input text;

[0016] The Transformer decoder is then used to decode the extracted features to achieve feature fusion.

[0017] In some embodiments, the method of concatenating the attribute text representations of all items in the same user interaction history sequence, performing feature extraction and fusion to obtain the text embedding representation of the corresponding user includes:

[0018] Concatenate the attribute text representations of all items in the order of recent to distant interaction time to form a long text sentence;

[0019] Splitting the long text sentence into a series of sub-paragraphs;

[0020] Using a Transformer encoder to extract features from each of the sub-paragraphs one by one to obtain corresponding sub-paragraph features;

[0021] Splicing the sub-paragraph features in the order of the sub-paragraphs to obtain splicing features of the long text sentence;

[0022] The concatenated features of the long text sentence are decoded using a Transformer decoder to obtain a text embedding representation of the corresponding user.

[0023] In some embodiments, the step of performing graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain a final graph embedding representation of the node includes:

[0024] The node aggregation calculation of graph convolution is implemented according to the following aggregation formula:

[0025]

[0026] In the formula, u is the node to be updated, is the set of neighbor nodes of node u, d u is the degree of node u, k is the number of aggregations, is the embedding representation of node u after k+1 aggregations, d i is the degree of node i, is the embedding representation of node i after k aggregations;

[0027] After a preset number of aggregations, the results after each aggregation are weighted summed to obtain the final graph embedding representation of the node.

[0028] In some embodiments, the step of fusing the text embedding representation of the target user and the final graph embedding representation to obtain the final embedding representation of the target user includes:

[0029] The text embedding representation and the final graph embedding representation of the target user are concatenated and sent to a multi-layer perceptron for linear transformation and nonlinear activation to obtain the final embedding representation of the target user.

[0030] In some embodiments, calculating the similarity between the final embedding representation of the target user and the text embedding representation of the item includes:

[0031] The similarity between the final embedding representation of the target user and the text embedding representation of the item is calculated using the vector inner product formula.

[0032] The present invention also provides an item debiasing recommendation system, which includes:

[0033] The text feature extraction module is used to obtain a user interaction history sequence and a user-item interaction graph, wherein the user interaction history sequence includes the user's historical interaction items, and the user-item interaction graph is an intersection adjacency matrix between users and items; the attribute text representations of all items in the same user interaction history sequence are concatenated and then subjected to feature extraction and fusion to obtain a text embedding representation of the corresponding user; the attribute text representations of the same item are subjected to feature extraction and fusion to obtain a text embedding representation of the corresponding item; the attribute text of the item includes popularity-related attributes and non-popularity-related attributes;

[0034] A graph contrast learning module is used to use the text embedding representation of the user and the text embedding representation of the item as the initial node embedding representation of the corresponding node in the user-item interaction graph; traverse all nodes in the user-item interaction graph, divide the directly connected nodes of the current node into two sets, aggregate the nodes in each set respectively, use one of the aggregation results as the embedding representation of the current node, use the other aggregation result as the positive sample of the current node, randomly select positive samples of multiple nodes from nodes other than the directly connected nodes as negative samples of the current node, and optimize the embedding representation of each node through contrast learning;

[0035] The graph convolution module is used to perform graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the node;

[0036] A feature fusion module is used to fuse the text embedding representation of the target user and the final graph embedding representation to obtain the final embedding representation of the target user;

[0037] The output module is used to calculate the similarity between the final embedding representation of the target user and the text embedding representation of the item, and determine the recommendation priority of the item based on the similarity. The higher the similarity, the higher the recommendation priority.

[0038] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of any of the above methods when executed by a processor.

[0039] The present invention also provides a computer program product, comprising a computer program or instructions, wherein when the computer program or instructions are executed by a processor, the steps of any of the above methods are implemented.

[0040] In general, compared with the prior art, the above technical solutions conceived by the present invention mainly have the following beneficial effects:

[0041] 1. The present invention integrates text embedding representation and graph embedding representation, uses the generalization of natural language to solve the popularity bias problem existing in traditional models, and accurately captures user interests through text matching to improve the model recommendation performance. Among them, when obtaining the text embedding representation, by splicing the attributes of all items in the user's historical interaction sequence and extracting information, the text attribute information is fully utilized, including both popularity-related attributes and non-popularity-related attributes, not limited to the specific attributes of the items, and has strong generalization. When obtaining the graph embedding representation, contrastive learning based on graph enhancement is adopted, and node positive and negative samples are constructed through node partitioning, so that more long-tail nodes can participate in feature extraction, which improves the recommendation frequency of long-tail nodes and solves the popularity bias problem caused by the long-tail distribution of data.

[0042] 2. Furthermore, in the embodiment of the present invention, the aggregation weight of the graph convolution is also optimized, which can suppress the information propagation of popular nodes, reduce popularity deviation, reduce the influence of popular nodes and non-popular nodes on other nodes, alleviate the center effect of the graph convolution network, and achieve a win-win situation of debiasing performance and recommendation performance.

[0043] 3. Furthermore, in the embodiment of the present invention, the user's long text sentences are segmented and features are extracted, and then the features are concatenated and fused to generate the user's text embedding representation, which can effectively reduce the time complexity of attention calculation and take into account the semantic features of different levels in the historical interaction sequence. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flowchart of the steps of a method for debiasing item recommendation in one embodiment of the present invention;

[0045] Figure 2 It is a schematic diagram of expanding some attributes of a certain commodity in one embodiment of the present invention;

[0046] Figure 3 is a schematic diagram of comparative learning of a certain block diagram in one embodiment of the present invention;

[0047] Figure 4 4 is a schematic diagram of the structure of an item debiasing recommendation system in one embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0049] Example 1

[0050] like Figure 1 The flowchart is a step flow chart of the method for debiasing recommendation of items in one embodiment of the present invention, which mainly includes steps S1 to S5. Each step is introduced in detail below.

[0051] Step S1: Obtain a user interaction history sequence and a user-item interaction graph, wherein the user interaction history sequence includes the user's historical interaction items, and the user-item interaction graph is an intersection adjacency matrix between users and items; concatenate the attribute text representations of all items in the same user interaction history sequence, perform feature extraction and fusion, and obtain a text embedding representation of the corresponding user; perform feature extraction and fusion on the attribute text representation of the same item, and obtain a text embedding representation of the corresponding item.

[0052] This step can be specifically performed as follows.

[0053] Step S11: Obtain user interaction history sequence.

[0054] Specifically, items and users whose number of interactions is lower than a preset requirement may be removed. For example, items and users whose number of interactions is too small may be removed by using the ten-degree screening method, and then a user interaction history sequence may be generated based on the interaction records.

[0055] Step S12: Traverse the user interaction history sequence and generate a user-item interaction adjacency matrix.

[0056] Specifically, both items and users can be used as nodes, for example, users can be used as column nodes, items can be used as column nodes, and the element in the Mth row and Nth column indicates whether the Mth user and the Nth item have an interactive relationship. The interactive adjacency matrix is ​​converted into a compression coefficient row for storage.

[0057] Step S13: Generate a text embedding representation for each user and a text embedding representation for each item.

[0058] For user nodes, the text representation of each user is obtained by concatenating the attribute text representations of all items in their historical interaction sequence. Each item has multiple attributes, including attributes related to popularity, such as evaluation scores, number of comments, etc., and attributes unrelated to popularity, such as item name, address, etc. In the present invention, the selected attributes include popularity-related attributes and non-popularity-related attributes. Before concatenating the attribute text representations of different items, multiple attribute key-value pairs of the same item are first expanded.

[0059] For item nodes, expand their multiple attributes to obtain the text representation of each item.

[0060] like Figure 2The figure shows a schematic diagram of expanding some attributes of a commodity in one embodiment of the present invention. For any item v, its attributes can be expanded into a sentence F(v), which is expressed as follows:

[0061] F(v)=Attr1:v(Attr1),…,Attr k :v(Attr k )

[0062] In the formula, Attr is the key of a specific attribute pair, and v(Attr) represents the value of the attribute pair.

[0063] In one embodiment, all attribute information of an item can be selected to implement full-text information extraction and obtain full-text representation embedding of the node. The advantage of using such full-text representation embedding is that it has no obvious tendency, and the embedding of long-tail items and popular items will not be affected by popularity bias. It can also make full use of the generalization advantages of natural language and the advantages of pre-trained models. Ultimately, through text matching between users and items, it can better capture the real interests of users and achieve a win-win situation of recommendation performance and debiasing performance.

[0064] Specifically, the Transformer encoder can be used to extract features from the text of each node, and then the Transformer decoder can be used to decode the extracted features to achieve feature fusion and obtain the text embedding representation of the corresponding node.

[0065] Considering that the user's text representation is long, it is difficult to extract fine-grained features and multi-layer semantics from a long text sequence. In order to improve the effect of feature extraction and fusion, in one embodiment, the following steps S131 to S135 are used to generate a text embedding representation for each user.

[0066] Step S131: Concatenate the attribute text representations of all items in the order of the interaction time of the items from recent to distant to form a long text sentence.

[0067] In order to understand the user's interaction history To encode, we first need to reverse the user's interaction history to In the form of t is the item that the current user interacted with at time t. Intuitively, the more recent the interaction record, the higher the weight. After reversing the interaction history, the text representation of the item can be used to replace the interaction history. , get the user's text representation, and finally, add an identifier to the front of each interaction record to indicate the beginning, as shown below:

[0068]

[0069] Finally, the text representation corresponding to each user contains the corresponding attributes of all items that the user has interacted with, which is equivalent to a long sentence composed of attribute key-value pairs of all items in the historical interaction sequence. Among the various attributes of items, some attributes are related to popularity, such as item ratings, total number of item reviews, etc. This part will also introduce the influence of popularity bias, but considering that this tendency is not always harmful, some popularity-related attributes are retained in the final text representation.

[0070] Step S132: Divide the long text sentence into a series of sub-paragraphs.

[0071] Specifically, split the long text sequence into a series of sub-paragraphs as follows:

[0072]

[0073] Here S n Represents the nth subsequence split into, each subsequence has m tokens.

[0074] Step S133: Use the Transformer encoder to extract features from each sub-paragraph one by one to obtain corresponding sub-paragraph features.

[0075] Specifically, the Transformer encoder is used for feature extraction, and the extracted features represent the user's interest preferences in a short period of time.

[0076] For example, for any subparagraph S i , perform feature extraction to obtain the sub-paragraph feature h i , which can be expressed as:

[0077] h i =TEncoder(S i )

[0078] Step S134: concatenate the sub-paragraph features in the order of the sub-paragraphs to obtain concatenated features of the long text sentence.

[0079] Concatenate the sub-paragraph features of the same user to represent the concatenated features of the long text sentence It is expressed as follows:

[0080]

[0081] Step S135: Use the Transformer decoder to decode the concatenated features of the long text sentence to obtain the text embedding representation of the corresponding user.

[0082] Join features By sending it to the Transformer decoder for decoding, we can get the user's complete interest preferences, as shown below:

[0083]

[0084] Generating the user's text embedding representation through the above steps can not only effectively reduce the time complexity of attention calculation, but also take into account the semantic features of different levels in the historical interaction sequence.

[0085] The generation method of the text embedding representation of an item is similar to that of the text embedding representation of a user. However, the attribute text representation of an item only has the attributes of a single item. The text is shorter, there is no concatenation operation, and it does not need to be divided into sub-paragraphs. The Transformer encoder and Transformer decoder are directly used for feature extraction and fusion to obtain the text embedding representation of the corresponding item.

[0086] Step S2: Use the text embedding representation of the user and the text embedding representation of the item as the initial node embedding representation of the corresponding node in the user-item interaction graph; traverse all nodes in the user-item interaction graph, divide the directly connected nodes of the current node into two sets, aggregate the nodes in each set respectively, use one of the aggregation results as the embedding representation of the current node, use the other aggregation result as the positive sample of the current node, randomly select multiple positive samples of nodes from nodes other than the directly connected nodes as negative samples of the current node, and optimize the embedding representation of each node through comparative learning.

[0087] Specifically, the user-item interaction graph can be divided into multiple block graphs, such as Figure 3 The figure shows a schematic diagram of a block diagram comparison learning in an embodiment of the present invention. For any node i, its directly connected nodes can be divided into a set and collection Respectively in the set and located in the set The directly connected nodes in the cluster are aggregated to obtain the embedding representation e of node i i and the positive sample e' of node i i , randomly select multiple nodes from nodes other than directly connected nodes and use them as positive samples as negative samples of node i, and optimize the embedding representation e of each node i through contrastive learning i .

[0088] The above process can be decomposed into the following operation steps S21 to S23.

[0089] Step S21: For a target node i, divide its directly connected nodes into two sets and Then, the neighbor nodes of the two sets are aggregated to obtain the embedded representation e of node i. i and positive sample e' i .

[0090] Specifically, for each node, its embedded representation and positive samples are obtained through aggregation.

[0091] For a target node i, its neighbor nodes are divided into two sets and Thus, the block graph to which it belongs is divided into two subgraphs, and then the embedding representation e of node i is obtained by aggregating the neighbor nodes of the two subgraphs i and e' i Here i and e' i is obtained from two completely different sets of neighbor sets, but the two neighbor sets are actually neighbors of node i in a block graph, so e i and e' i should be similar, so e' i As e i A positive sample of .

[0092] Step S22: randomly select multiple nodes from the remaining non-directly connected nodes and use their positive samples as negative samples of the target node i. Here, not every negative sample in each batch is involved in the calculation, but a part of the nodes are randomly selected from the negative sample set to reduce the calculation complexity.

[0093] Step S23: Perform comparative learning using the InfoNCE evaluator.

[0094] The loss function of contrastive learning is:

[0095]

[0096] Where B is the node set in the block graph divided by the user-item interaction graph, and B i is the set of nodes selected as negative samples of node i in the block graph, sim(e i ,e' i ) is to calculate the embedding representation e i and positive sample e' i The similarity function, sim(e i ,e' j ) is to calculate the embedding representation e i and negative samples e' j Here, cosine similarity can be used as a measure of vector similarity, which can avoid the scale effect and zero vector problems caused by direct dot product.

[0097] The optimization method of contrastive learning is to treat the node representation as a random variable and maximize the mutual information between variables. Each node is aggregated to obtain e i and e' i Two expressions, where e' i is the positive sample of the node. The node represents the positive sample that should be closer to itself, and the nodes other than the neighboring nodes are the targets that should be kept away from. Therefore, e' i As positive samples, the positive samples e' of nodes other than neighbor nodes are taken j As its negative sample, and optimize according to the comparison of positive and negative samples. In the present invention, through comparative learning, the long-tail nodes can be involved in the training to reduce the popularity deviation.

[0098] Step S3: Perform graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the node.

[0099] For each node, the features of its neighboring nodes are aggregated through graph convolution to learn the relationship between nodes, thereby achieving feature extraction and classification.

[0100] Furthermore, in order to suppress the spillover effect of popularity nodes on other nodes, in one embodiment, the aggregation weight of graph convolution is designed. For any target node u, when its neighbor nodes are aggregated to obtain the representation of the target node u, the weight is used. As the weight of neighbor node i, d u is the degree of the target node u (that is, the number of neighboring nodes), and the specific aggregation formula is as follows:

[0101]

[0102] In the formula, u is the node to be updated, is the set of neighbor nodes of node u, d u is the degree of node u, k is the number of aggregations, is the embedding representation of node u after k+1 aggregations, d i is the degree of node i, is the embedding representation of node i after k aggregations.

[0103] After a preset number of aggregations, the results after each aggregation are weighted summed to obtain the final graph embedding representation of the node.

[0104] From the perspective of information propagation, if a node i can reach node j after passing through several other nodes, that is, there is a path from i to j in the graph, then the representation of node i will have an impact on node j after propagation. If node i can reach node j, then from a mathematical expectation, the number of paths should be related to the degree of node i. The more the degree, the more expected the number of paths. Here, the influence of one node on another node in the graph is evaluated by the derivative of the BPR loss function, and the results are as follows:

[0105]

[0106]

[0107] HereΦ u represents the set of paths of length L-1 from node u to node j, and Φ is the set of paths of length L from node i to node j. u The change will be different by a degree d of i i , in terms of expectations So we can extract a d from the formula i , and then divided by the normalized Here L is the node i, and we can get the value in the second line of the formula Finally, use δ i Replacing the expected value in the formula can be more concise.

[0108] It can be found that the influence of node i on node j is related to its degree, and the larger the degree, the greater its influence The larger the value, the larger the value. Therefore, based on this conclusion, the weight of node aggregation can be dynamically adjusted. When the target node u is aggregated with neighboring nodes to obtain a representation, for each neighboring node i, use As its weight, this can effectively suppress the spillover effect of popularity nodes on other nodes. Through the improved aggregator, the gap in the influence of popular nodes and non-popular nodes on other nodes is narrowed, and the trend of other nodes approaching popular nodes is slowed down, which can largely solve the problem of central effect and effectively eliminate the popularity bias in the system.

[0109] Step S4: Feature fusion is performed on the text embedding representation of the target user and the final graph embedding representation to obtain the final embedding representation of the target user.

[0110] Among them, the target user is the user to whom items need to be recommended.

[0111] Specifically, the text embedding representation and the final graph embedding representation of the target user can be concatenated and sent to a multi-layer perceptron for linear transformation and nonlinear activation to obtain the final embedding representation of the target user.

[0112] The main function of feature fusion is to fuse different embedding representations to obtain the final user representation, which is used to recommend candidate item sets for the target user. For a given user historical interaction sequence, text information is extracted to obtain its text-based text embedding representation T u , and then the graph embedding representation I is obtained by extracting the interaction information from the user-item interaction graph through the graph convolutional network u , concatenate the two vectors:

[0113] X u =Concat(T u ,I u )

[0114] The concatenated vector is input into the multi-layer perceptron, and after linear transformation and nonlinear activation, the final embedding representation of the target user can be obtained: u =MLP(X u ).

[0115] Step S5: Calculate the similarity between the final embedding representation of the target user and the text embedding representation of the item, and determine the recommendation priority of the item according to the similarity. The higher the similarity, the higher the recommendation priority.

[0116] Specifically, the similarity between the final embedding representation of the target user and the text embedding representation of the item can be calculated by the vector inner product formula. The calculation formula is as follows:

[0117]

[0118] Where P w is the text embedding representation of item w in the candidate set, is the similarity score between the final embedding representation of the target user and the text embedding representation of item w, and is also the ranking score in the final recommendation candidate set. The top-k items with the highest scores are selected as the final output and recommended to the user.

[0119] In general, the technical solution of the embodiment of the present invention, by integrating text embedding representation and graph structure embedding representation and introducing the advantages of comparative learning models, comprehensively analyzes the user's interest data, generates a recommendation list of interest points that meets the user's preferences, removes the popularity bias in the recommendation system, reduces the proportion of popular items in the recommendation results, recommends items that truly meet the user's interests, and optimizes the user's usage experience.

[0120] Example 2

[0121] The present invention also relates to an item debiasing recommendation system, such as Figure 4FIG. 1 is a schematic diagram of the structure of an item debiasing recommendation system in an embodiment of the present invention, which includes:

[0122] The text feature extraction module is used to obtain the user interaction history sequence and the user-item interaction graph. The user interaction history sequence contains the user's historical interaction items. The user-item interaction graph is the user-item interaction adjacency matrix. The attribute text representations of all items in the same user interaction history sequence are concatenated and then feature extracted and fused to obtain the text embedding representation of the corresponding user. The attribute text representations of the same item are feature extracted and fused to obtain the text embedding representation of the corresponding item. The attribute text of the item contains popularity-related attributes and non-popularity-related attributes.

[0123] A graph contrast learning module is used to use the text embedding representation of the user and the text embedding representation of the item as the initial node embedding representation of the corresponding node in the user-item interaction graph; traverse all nodes in the user-item interaction graph, divide the directly connected nodes of the current node into two sets, aggregate the nodes in each set respectively, use one of the aggregation results as the embedding representation of the current node, use the other aggregation result as the positive sample of the current node, randomly select multiple positive samples of nodes from nodes other than the directly connected nodes as negative samples of the current node, and optimize the embedding representation of each node through contrast learning;

[0124] The graph convolution module is used to perform graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the node;

[0125] A feature fusion module is used to fuse the text embedding representation of the target user and the final graph embedding representation to obtain the final embedding representation of the target user;

[0126] The output module is used to calculate the similarity between the final embedding representation of the target user and the text embedding representation of the item, and determine the recommendation priority of the item based on the similarity. The higher the similarity, the higher the recommendation priority.

[0127] It can be understood that the system in Example 2 can be used to execute the method in Example 1. For specific details, please refer to the introduction of Example 1 and will not be repeated here.

[0128] Example 3

[0129] The present invention also relates to a computer-readable storage medium having a computer program stored thereon, and the computer program implements the steps of the above method when executed by a processor.

[0130] Specifically, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, an internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0131] Example 4

[0132] The embodiment of the present invention provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps of the method of the above embodiment of the present invention.

[0133] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. It should be noted that "in one embodiment", "for example", "for example", etc. of the present invention are intended to illustrate the present invention, rather than to limit the present invention.

[0134] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent application. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention.

Claims

1. A method for debiasing item recommendation, characterized in that: include: Obtain a user interaction history sequence and a user-item interaction graph, wherein the user interaction history sequence includes the user's historical interaction items, and the user-item interaction graph is an intersection adjacency matrix of users and items; concatenate the attribute text representations of all items in the same user interaction history sequence, perform feature extraction and fusion, and obtain a text embedding representation of the corresponding user; perform feature extraction and fusion on the attribute text representation of the same item, and obtain a text embedding representation of the corresponding item; the attribute text of the item includes popularity-related attributes and non-popularity-related attributes; Using the text embedding representation of the user and the text embedding representation of the item as initial node embedding representations of corresponding nodes in the user-item interaction graph; Traversing all nodes in the user-item interaction graph, dividing directly connected nodes of the current node into two sets, respectively aggregating the embedded representations of the nodes in each set, using one of the aggregated results as the embedded representation of the current node, using the other aggregated result as the positive sample of the current node, randomly selecting positive samples of multiple nodes from nodes other than directly connected nodes as negative samples of the current node, and optimizing the embedded representation of each node through comparative learning; Perform graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the nodes; The text embedding representation of the target user and the final graph embedding representation are feature-fused to obtain the final embedding representation of the target user; Calculate the similarity between the final embedding representation of the target user and the text embedding representation of the item, and determine the recommendation priority of the item based on the similarity. The higher the similarity, the higher the recommendation priority.

2. The method for debiasing item recommendation according to claim 1, characterized in that: The obtaining of the user interaction history sequence and the user-item interaction graph includes: After removing items and users whose interaction numbers are lower than a preset requirement, the user interaction history sequence and the user-item interaction graph are generated according to the interaction records.

3. The method for debiasing item recommendation according to claim 1, wherein: The feature extraction and fusion comprises: First, use the Transformer encoder to extract features from the input text; The Transformer decoder is then used to decode the extracted features to achieve feature fusion.

4. The method for debiasing item recommendation according to claim 3, characterized in that: The method of concatenating the attribute text representations of all items in the same user interaction history sequence, performing feature extraction and fusion, and obtaining the text embedding representation of the corresponding user includes: Concatenate the attribute text representations of all items in the order of recent to distant interaction time to form a long text sentence; Splitting the long text sentence into a series of sub-paragraphs; Using a Transformer encoder to extract features from each of the sub-paragraphs one by one to obtain corresponding sub-paragraph features; Splicing the sub-paragraph features in the order of the sub-paragraphs to obtain splicing features of the long text sentence; The concatenated features of the long text sentence are decoded using a Transformer decoder to obtain a text embedding representation of the corresponding user.

5. The method for debiasing item recommendation according to claim 1, wherein: The method of performing graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the node includes: The node aggregation calculation of graph convolution is implemented according to the following aggregation formula: In the formula, u is the node to be updated, is the set of neighbor nodes of node u, d u is the degree of node u, k is the number of aggregations, is the embedding representation of node u after k+1 aggregations, d i is the degree of node i, is the embedding representation of node i after k aggregations; After performing a preset number of aggregations, the results after each aggregation are weighted summed to obtain the final graph embedding representation of the node.

6. The method for debiasing item recommendation according to claim 1, wherein: The step of fusing the text embedding representation of the target user and the final graph embedding representation to obtain the final embedding representation of the target user includes: The text embedding representation and the final graph embedding representation of the target user are concatenated and sent to a multi-layer perceptron for linear transformation and nonlinear activation to obtain the final embedding representation of the target user.

7. The method for debiasing item recommendation according to claim 1, wherein: The calculating the similarity between the final embedding representation of the target user and the text embedding representation of the item includes: The similarity between the final embedding representation of the target user and the text embedding representation of the item is calculated using the vector inner product formula.

8. An item debiasing recommendation system, characterized in that: include: The text feature extraction module is used to obtain a user interaction history sequence and a user-item interaction graph, wherein the user interaction history sequence includes the user's historical interaction items, and the user-item interaction graph is an intersection adjacency matrix between users and items; the attribute text representations of all items in the same user interaction history sequence are concatenated and then subjected to feature extraction and fusion to obtain a text embedding representation of the corresponding user; the attribute text representations of the same item are subjected to feature extraction and fusion to obtain a text embedding representation of the corresponding item; the attribute text of the item includes popularity-related attributes and non-popularity-related attributes; A graph comparison learning module, configured to use a text embedding representation of a user and a text embedding representation of an item as initial node embedding representations of corresponding nodes in the user-item interaction graph; Traversing all nodes in the user-item interaction graph, dividing directly connected nodes of the current node into two sets, respectively aggregating the embedded representations of the nodes in each set, using one of the aggregated results as the embedded representation of the current node, using the other aggregated result as the positive sample of the current node, randomly selecting positive samples of multiple nodes from nodes other than directly connected nodes as negative samples of the current node, and optimizing the embedded representation of each node through comparative learning; The graph convolution module is used to perform graph convolution feature extraction on the user-item interaction graph after contrastive learning to obtain the final graph embedding representation of the node; A feature fusion module is used to fuse the text embedding representation of the target user and the final graph embedding representation to obtain the final embedding representation of the target user; The output module is used to calculate the similarity between the final embedding representation of the target user and the text embedding representation of the item, and determine the recommendation priority of the item based on the similarity. The higher the similarity, the higher the recommendation priority.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images